Function Optimization and Benchmarking Based on Six Wolf Algorithms

Optimization algorithms are known in science field due to their fast speed in finding the solution and consistency in many optimization problems.   As of recent, one mammal was used many times as a inspiration for those mechanism – wolf.  However, since there are many of them, there isn’t a clear view of which one is better then the other.  Six wolf pack algorithms are described and checked in this paper.  R was chosen as a language to implement them, as well as six benchmarks.  In order to gather data, algorithms worked thirty times on each of functions.  Results are presented as a mean score, standard deviation of the score, mean time and standard deviation of the time.  Furthermore, convergence plot is presented from the hardest benchmark.  The wolf algorithms were compared to PSO Kennedy & Eberhart (1995), DE Storn & Price (1997), and GA Holland (1992). The winner of this comparison is Grey Wolf Optimizer Mirjalili et al. (2014).